ORIGINAL
Painless thyroiditis mimicking relapse of hyperthyroidism during or after potassium iodide or thionamide therapy for Graves’ disease resulting in remission
2023 年 70 巻 2 号 p. 207-222
詳細
2023 年 70 巻 2 号 p. 207-222
The diagnosis of painless thyroiditis (PT) during antithyroid drug (ATD) treatment of Graves’ disease (GD) is difficult. We evaluated the thyroidal radioactive iodine uptake (RAIU) in 100 patients with relapsed thyrotoxicosis during or after careful ATD treatment. The RAIU was <5%/5 h in 35 patients (35%) (Group A - PT), 5%–15%/5 h in 6 patients (6%) (Group B - indefinite) and >15%/5 h in 59 patients (59%) (Group C - relapsed GD [rGD]). TSH receptor antibody (TBII) was positive in 4 (11.4%), 3 (50.0%) and 39 (only 66.1%) patients in Groups A, B and C, respectively. In Group A, the serum fT4 level spontaneously normalized after 35 (26–56) days, sometimes followed by transient hypothyroidism, confirming the diagnosis of PT. Nineteen (54.3%) had been treated with potassium iodide, and PT frequently occurred ironically when the ATD dosage was reduced. PT repeatedly occurred in nine patients. All went into remission smoothly or developed hypothyroidism, except one patient with strongly positive TBII who developed rGD after the resolution of PT (PT on GD). In 10 (50%) of 20 patients with negative TBII despite rGD in Group C, TBII became positive afterwards. In conclusion, it is important to recognize that PT can occur in the clinical course of GD, resulting in frequent remission despite relapse of PT. The thyroid function reflects the balance between the stimulating TBII activity and the responsiveness of the thyroid tissue (sometimes unresponsive and other times autostimulated). The RAIU is still a valuable tool in cases of ambiguous thyrotoxicosis.
GRAVES’ DISEASE (GD) and Hashimoto thyroiditis are recognized as being pathologically interrelated, as GD may occur in patients whose thyroid glands histologically show either Hashimoto thyroiditis alone or a mixture of both parenchymatous hypertrophy of GD and extensive lymphocytic infiltration [1]. These two conditions may represent a single disease entity with a wide range of manifestations. The concept of painless thyroiditis (painless low-uptake thyrotoxicosis without hyperthyroidism; PT) has been recognized since 1975 [2-10], and postpartum PT has been reported since 1977 [11, 12]. Histologically, chronic thyroiditis, focal or diffuse type, is shown, and lymphoid follicles are present in about a half of patients, without stromal fibrosis or oxyphilic cell changes [10]. Follicular disruption is a characteristic feature of this disorder and disappears after recovery. Giant cells are reported in about two thirds of specimens.
The differential diagnosis of PT and GD is not difficult before treatment. The golden-standard factor to consider for the differential diagnosis is the thyroidal radioactive iodine uptake (RAIU), which is high in GD and almost null in typical PT cases. PT is also suggested when thyroid-stimulating hormone (TSH) receptor antibody (TRAb), measured by the TSH Binding Inhibitor Immunoglobulin (TBII) or thyroid-stimulating antibody activity (TSAb), is negative and thyrotoxicosis resolves spontaneously without antithyroid drugs (ATDs) followed by an episode of transient hypothyroidism.
Postpartum PT after delivery in patients with GD has been reported since 1980 [13-16]. The onset of GD after an episode of PT has also been reported [17, 18]. Ozawa clinically suggested a close relationship between PT and GD, reporting that 7 of 36 PT patients had a history of GD [19]; this suggested that PT and GD may clinically represent different aspects of a single disease entity. Therefore, the diagnosis is difficult when PT occurs during ATD treatment of GD. Misaki reported the usefulness of the Technetium-99m uptake for diagnosing PT during the clinical course of GD [20].
Recently, PT has drawn attention because of the high incidence of PT following immune checkpoint inhibitor treatment [21-28]. We herein report an illustrative case and a high incidence of PT during or after potassium iodide (KI) or thionamide treatment of GD, resulting in an increased chance of remission. We also describe a case in which TBII and TSAb were negative at the apparent relapse of GD with a high RAIU.
From 1985 to 2005, the RAIU was determined in 100 patients with unexpected thyrotoxicosis (elevation of serum free T4 [fT4] and/or free T3 [fT3] levels) that was observed during or after the careful treatment of GD [29]. The patients’ poor medication compliance excluded the possibility of unexpected thyrotoxicosis. The patients had been treated for more than 180 days with an ATD, such as methylmercaptoimidazole (MMI), propylthiouracil (PTU) [30, 31] or KI [32, 33]. If thyrotoxicosis resolved without increasing the dosage of ATD and/or followed by hypothyroidism, PT was suggested, and the diagnosis was confirmed when the RAIU measured at the time of relapse was <5%/5 h (Group A) [34]. If the RAIU was between 5%–15%/5 h, the episodes were diagnosed as indefinite type (Group B). If the RAIU was >15%/5 h requiring an increase in the ATD dose, relapse of GD was diagnosed (Group C).
The clinical features of 105 patients with untreated PT and 1,124 patients with untreated GD who visited our hospital during the same period were compared.
The serum fT3, fT4, TSH, autoantibodies to thyroglobulin (Tg) (TGHA) and thyroid microsomal antigen (MCHA) were measured as previously reported [33, 35]. The serum TBII level was mostly measured using a first-generation radioreceptor assay kit (Baxter Health Care Co. Ltd., Tokyo, Japan; normal range <15%). Starting in 2004, during long-term follow up, TBII was also measured using a second-generation TBII assay with the human recombinant TSH receptor (normal <1 IU/L, DYNOtest TRAb human kit; Yamasa Corporation, Chiba, Japan). TSAb measuring c-AMP produced in FRTL-5 cells, and the estimated thyroid volume were measured as previously reported [33, 35]. The RAIU was measured before treatment of GD (at first visit) and at the relapse of thyrotoxicosis during or after ATD therapy [34]. The patients had been asked to refrain from eating iodine-rich foods, such as seaweed and tangle, and to stop taking ATD at the relapse, for at least seven days before the test. None of the patients had received X-ray contrast medium within six months before the RAIU measurement. After the introduction of ultrasonography and 123I instead of 131I, the RAIU was measured five hours after receiving a minimum dose of 123I without scintigraphy in patients with diffuse goiter in order to minimize the medical cost and hospital visits for the patient [34]. Regarding the concern about the influence of KI therapy on the RAIU test, most of the excess iodide that did not enter the thyroid gland was rapidly excreted into the urine and KI therapy did not interfere with the RAIU test after iodine restriction for a week [32, 33]. Although the normal value for the RAIU is 5%–20%/5 h in our laboratory [34], relapse of GD was diagnosed when the RAIU was >15%/5 h, considering the suppressed TSH level in thyrotoxicosis. The RAIU is 2%–7%/5 h in cases of pituitary insufficiency and 0%–5%/5 h in cases of untreated PT in our laboratory [34].
The reference values in our laboratory were as follows: serum fT4 0.8–1.7 ng/dL, fT3 2.2–3.8 pg/mL, TSH 0.42–3.81 mU/L, TBII <15%, TSAb <180%, the RAIU 5%–20%/5 h, thyroid volume 5 mL–20 mL, TGHA or MCHA measured by hemagglutination <100 dilution. If patients maintained a euthyroid status with normal serum TSH levels and negative TBII for over one year after the cessation of the drug, they were considered to have entered remission.
Statistical analysesNormal continuous variables were expressed as the mean ± standard deviation, while median and interquartile ranges were reported for skewed variables. Comparisons were performed using Pearson’s χ2 test between categories and Student’s t-test, Wilcoxon’s rank sum test or a multivariate logistic regression analysis between continuous variables. The regression was calculated using the step-wise method. The lengths of time required to achieve remission were compared using the log-rank test. The analyses were performed using the JMP 16 software program (SAS Institute, Inc., Cary, NC, USA). A p value below 0.05 was considered to be statistically significant.
The study was approved by the Ethics Committee of Kyushu University. This article does not describe any studies with human or animal subjects performed by any of the authors. The identity of the patient has been protected. The subjects were allowed to opt out of the present study. Patient 1 provided her written consent for inclusion of her data in this manuscript.
The first case of PT observed during ATD treatment of GD is shown (Fig. 1).
Changes in the serum fT4 and TSH levels in a patient with Graves’ disease (GD) with repeated episodes of thyrotoxicosis during MMI or KI treatment (Patient 1). The 2nd and 3rd episodes seemed to be relapse of GD, as suggested by strong TBII and TSAb but the 4th–8th episodes seemed to be painless thyroiditis (PT), as confirmed by the low RAIU in the 7th and 8th episodes. The patient eventually became spontaneously hypothyroid and was treated with synthesized L-T4. MMI: methylmercaptoimidazole, KI: potassium iodide, L-T4: L-thyroxine, TBII: TSH binding inhibitor immunoglobulin, TSAb: thyroid-stimulating antibody, RAIU: thyroidal radioactive iodine uptake. Abnormally high values are indicated in red. Thyroid vol: Estimated thyroid volume. A large thyroid volume (≥30 mL) is underlined, and an extremely large thyroid volume (≥70 mL) is indicated in red.
A 27-year-old woman with GD was treated with 30 mg MMI (1st episode). The RAIU was 78.6%/5 h before treatment. She went into remission after two years. At 40 years old, relapse of GD was observed when the serum fT4 level was 4.7 ng/dL, fT3 11.6 pg/mL, TSH <0.01 mU/L, TBII 26.3%, and the RAIU 46.4%/5 h (2nd episode). She was treated with 15 mg MMI, the dosage of which was later reduced with a maintenance dose (5 mg) continued for 4 years. Four months after the MMI dosage was reduced to 5 mg/3 days, re-elevation of the serum fT4 level (7.5 ng/dL) was observed (3rd episode). TBII (42.6%) and TSAb (1,360%) were both strongly positive, and the MMI dosage was increased to 10 mg. The serum fT4 level became very low (0.5 ng/dL), so the MMI dosage was reduced to 5 mg, and the serum fT4 level then normalized (1.0 ng/dL).
After this episode, MMI was changed to KI (30 mg/day). After 5 months, the serum fT4 level increased to 3.3 ng/dL (4th episode), so the KI dosage was increased to 100 mg. After 3 months, she became severely hypothyroid with undetectable serum fT4 levels (<0.1 ng/dL), extremely elevated serum TSH levels (82 mU/L) and an enlarged thyroid gland (70 mL). KI was withdrawn for 14 days. She was then treated with 50 mg KI and became euthyroid.
Twelve months later, after experiencing a stressful personal event, her serum fT4 level increased to 6.8 ng/dL (5th episode), so the KI dosage was increased to 200 mg. She again became severely hypothyroid 3 months later with an extremely enlarged thyroid gland (121 mL) and treated with 50 mg KI and 25–100 μg synthesized L-thyroxine (L-T4). One year after the 5th episode of thyrotoxicosis, the serum fT4 level again increased to 3.4 ng/dL (6th episode), and she was treated with 100 mg KI and 100 μg L-T4. A hypothyroid phase was not observed after the 6th episode, and the TSH level remained suppressed on L-T4.
About 16 months later, the serum fT4 level slightly increased to 2.4 ng/dL but she remained asymptomatic. A re-examination 4 months later revealed a markedly high serum fT4 level (4.7 ng/dL) with complaints of shortness of breath, enlarged goiter and brisk Achilles tendon reflex. TSAb was slightly positive. At this 7th episode of relapse, all drugs were withdrawn, and the RAIU was re-examined, showing a value of 1.6%/5 h and confirming the diagnosis of PT.
She became slightly hypothyroid without any drugs (fT4 0.5 ng/dL, TSH 8.3 mU/L) and then spontaneously became euthyroid despite marginally elevated TBII and TSAb levels. She became hypothyroid again about 2 years later (fT4 0.5 ng/dL, TSH 35.1 mU/L) and was treated with 50–75 μg L-T4. She remained well without any ATDs but with L-T4 for three years. The 8th episode of thyrotoxicosis (fT4 3.3 ng/dL) was observed at 54 years old due to the stress of administering nursing care to her aged father; the RAIU was again as low as 0.8%/5 h, suggesting PT. She became hypothyroid 2 months later and was treated with 50 μg L-T4 until 70 years old without any further relapse of thyrotoxicosis.
The estimated thyroid volume was 10–30 mL in a stable condition, 30–60 mL in active GD or PT and 70–120 mL in the hypothyroid state while extremely stimulated by endogenous TSH. The estimated final thyroid volume at 70 years old was 17 mL with slightly heterogeneous echogenicity.
A similar clinical course of PT observed during the treatment of GD with KI or MMI is shown in Fig. 2. PT was usually observed when the dosage of KI or MMI was reduced, and elevation of TSH levels was observed when the dosage was increased. The interval between the thyrotoxic and hypothyroid phases, when present, was about 3 months, although the magnitude of the fT4 elevation and TSH elevation was varied.
Repeated episodes of thyrotoxicosis during MMI or KI treatment in Graves’ hyperthyroidism (GD). See the legend of Fig. 1. After the treatment of GD, as shown by the numbers in red, repeated relapse of thyrotoxicosis was observed, as shown by the numbers in green. A low thyroidal uptake of radioactive iodine confirmed the diagnosis of painless thyroiditis, as shown by the numbers in blue, and spontaneous recovery of the thyroid function was observed without antithyroid drugs.
Among 100 patients with relapse of thyrotoxicosis during or after careful ATD treatment for GD who had their RAIU measured, the RAIU was 0%–5%/5 h (suggesting PT) in 35 (35.0%) patients (Group A), >15% (suggesting relapse of GD) in 59 (59.0%) patients (Group C) and 5%–15% (suggesting indefinite type thyrotoxicosis) in 6 (6.0%) patients (Group B) (Table 1). Compared with the results in untreated GD (n = 1,124) or untreated PT (n = 105) observed in our hospital during the same period, the RAIU was suppressed in most patients with PT and >15% in most patients with GD in both series (Fig. 3). There were a few indefinite cases among relapsed thyrotoxicosis patients. The RAIU was 35.4% (23.7%–57.0%)/5 h in the relapsed GT group, which was significantly lower than the value of 55.2% (36.9%–70.5%)/5 h found in untreated GD (p < 0.0001).
| Number (%) | A) GD→PT 35 (35.0%) |
B) GD→Indefinite 6 (6.0%) |
C) GD→rGD 59 (59.0%) |
p value A vs. C |
|
|---|---|---|---|---|---|
| I) Before treatment | GD | GD | GD | ||
| Age (years) | 31.9 ± 13.3 | 34.0 ± 13.4 | 37.2 ± 14.0 | 0.0379 | |
| Sex (Male:Female) | 7:28 | 0:6 | 10:49 | 0.9578 | |
| fT4 (ng/dL) | 5.4 (3.6–8.7) | 4.4 (3.7–4.7) | 6.0 (3.6–7.9) | 0.1430 | |
| fT3 (pg/mL) | 13.3 (10.0–20.1) | 11.9 (10.1–17.1) | 13.5 (9.0–18.9) | 0.2796 | |
| fT3/fT4 (pg/mL/ng/dL) | 2.3 (2.1–2.9) | 2.7 (2.2–2.9) | 2.5 (1.8–3.3) | 0.9247 | |
| TBII (%) | 33.4 (17.6–57.2) | 26.9 (24.8–37.7) | 35.1 (19.1–52.4) | 0.9609 | |
| TSAb (%) | 251 (156–464) | 137 (122–165) | 215 (135–385) | 0.3372 | |
| TGHA (positive) | 25 (71.4%) | 2 (33.3%) | 26 (44.1%) | 0.0311 | |
| MCHA (positive) | 28 (80.0%) | 4 (66.7%) | 50 (84.7%) | 0.2178 | |
| Thyroid volume (mL) | 30 (20–36) | 25 (18–32) | 26 (19–40) | 0.5291 | |
| RAIU (%/5 h) | 52.5 (37.7–69.4) | 43.5 (31.6–61.1) | 52.9 (37.1–71.8) | 0.3899 | |
| II) Relapse of thyrotoxicosis | PT | Indefinite | rGD | ||
| Interval (days) | 1,875 (973–3,289) | 1,393 (693–2,130) | 1,407 (1,084–2,299) | 0.9671 | |
| Age (years) | 38.6 ± 16.9 | 37.3 ± 14.9 | 42.5 ± 13.8 | 0.6800 | |
| fT4 (ng/dL) | 3.3 (2.4–4.6) | 2.4 (1.9–2.5) | 3.1 (2.1–4.3) | 0.0121 | |
| fT3 (pg/mL) | 8.1 (5.6–12.0) | 7.5 (7.3–10.6) | 7.5 (5.7–12.8) | 0.0081 | |
| fT3/fT4 (pg/mL/ng/dL) | 2.5 (2.0–2.9) | 3.0 (3.0–3.2) | 3.0 (2.5–3.6) | 0.0017 | |
| Thyroid volume (mL) | 23 (13–30) | 15 (10–28) | 19 (10–31) | 0.8283 | |
| RAIU (%/5 h) | 2.4 (0.9–3.2) | 13.2 (10.9–14.1) | 35.4 (23.7–57.0) | ||
| TBII (%) | 4.8 (1.4–11.1) | 12.1 (2.1–39.7) | 21.8 (10.7–41.6) | ||
| TSAb (%) | 120 (94–146) | 132 (117–139) | 133 (109–260) | ||
| Postpartum | 9/20 (45.0%) | 1/3 (33.3%) | 6/22 (27.3%) | ||
GD: Graves’ hyperthyroidism, PT: Painless thyroiditis, rGD: relapse of GD, TBII: TSH binding inhibitor immunoglobulin, TSAb: thyroid-stimulating antibody, TGHA, MCHA: anti-thyroglobulin and andi-thyroid microsomal antibody measured by hemagglutination, RAIU: thyroidal radioactive iodine uptake, Interval: between the date of RAIU before treatment of GD and RAIU at the relapse of thyrotoxicosis. Postpartum: in female patients 20–40 years old at the relapse of thyrotoxicosis. The serum fT4 and fT3 levels at the relapse of GD (Group C-II) were significantly lower than those before treatment (Group C-I) (p = 0.0013 and p = 0.0023 , respectively). TSAb was positive in 54 (59.3%) of 91 patients measured before treatment of GD, and 2 (5.7%) in Group A, 1 (16.7%) in Group B and 21 (35.6%) in Group C at relapse of thyrotoxicosis. The prevalence of TBII positive patients is shown in Table 6.
The prevalence of the patients with high or suppressed thyroidal radioactive iodine uptake (RAIU) in patients with painless thyroiditis (PT) (green) or relapsed Graves’ hyperthyroidism (GD) (purple) during the treatment of GD, respectively, compared with the prevalence in untreated patients with PT (blue) or GD (red) in our hospital. In the patients in whom the RAIU was 5%–15%/5 h (Group B), the diagnosis was suggested by the clinical course after the episode of thyrotoxicosis (See text). * The RAIU test was performed 34 days after the relapsed thyrotoxicosis (serum TSH 0.01 mU/L, fT4 2.5 ng/dL, TBII 7.9%, TSAb 137%), when the patient was at the recovery phase of PT without antithyroid drug (serum TSH 0.22 mU/L, fT4 1.2 ng/dL, RAIU 7.2%/5 h) resulting in smooth remission of GD.
At the diagnosis of GD before treatment, the RAIU was high in all patients (Fig. 4a), and TBII was positive in 82 (82.0%) patients (Fig. 4b). Therefore, about 20% of the patients with untreated GD were TBII-negative with this first-generation assay. TSAb was positive in 54 (59.3%) of 91 patients measured. At relapse of thyrotoxicosis, as shown in Fig. 4b, TBII was positive in 4 (11.4%) patients in Group A and in 39 (66.1%) in Group C. The prevalence of TBII-negative patients with a relapse of GD was thus 33.9%, which was significantly higher than that before treatment (18 of 100 patients; 18.0%) (Fig. 4b) (p = 0.0232). TBII was positive in 3 (50.0%) patients in Group B. TSAb was positive in 24 (24.0%) of 100 patients at relapse and 21 (87.5%) of them were Group C. Therefore, TSAb was positive in 35.6% of 59 patients in Group C. The 24-h RAIU, evaluated in only some of the patients, was 62.6% (51.0%–70.0%) before treatment of GD (n = 74) and 2.2% (1.7%–2.7%) in Group A (n = 4) and 47.1% (30.1%–62.0%) in Group C (n = 16) at the relapse of thyrotoxicosis.
The thyroidal radioactive iodine uptake (5 h) (RAIU) and TSH binding inhibitor immunoglobulin activity (TBII) before treatment of Graves’ hyperthyroidism (GD) and at the time of the relapse of thyrotoxicosis. Relapse was diagnosed as painless thyroiditis (PT) in Group A, indefinite type (ID) in Group B and relapse of Graves’ hyperthyroidism (rGD) in Group C. TBII was positive in 82 (82.0%) of 100 patients before treatment of GD and only in 39 (66.1%) of 59 patients at the relapse of GD. The difference was significant (p = 0.0232). The reference range is indicated with an arrow and that for the RAIU is for TSH-suppressed GD patients described in the text.
In Group A, repeated episodes of PT were suggested to occur in 3 patients (6 times), 1 patient (5 times), 3 patients (3 times) and 2 patients (2 times), with these episodes initially misdiagnosed as relapse of GD (Figs. 1 and 2).
A comparison of Groups A and C suggested that there were no significant differences between groups before treatment, except for the lower age and higher incidence of patients with positive TGHA in Group A (Table 1-I). The relapse of thyrotoxicosis was confirmed by the RAIU after 1,875 days (5.1 years) as PT in Group A and after 1,407 days (3.9 years) as the relapse of GD in Group C after the initiation of treatment for GD (Table 1-II). At the time of relapse of thyrotoxicosis, the serum fT4 and fT3 levels were higher with a lower fT3/fT4 ratio in Group A than in other groups, without a significant difference in the thyroid volume. The RAIU and TBII values were markedly higher in Group C than in other groups, as might be expected. The prevalence of patients who showed relapse of thyrotoxicosis within 1 year postpartum (among female patients 20–40 years old), was 45.0% in Group A at 3.0 (2.0–4.0) months postpartum and 27.3% in Group C at 5.5 (4.3–6.8) months postpartum (Table 1-II).
Patients with indefinite-type thyrotoxicosis (Group B)There were six female patients in Group B and TBII was positive in three and negative in three (Fig. 4b). Among the TBII-negative patients, 1 patient (fT4 2.5 ng/dL, TSH <0.01 mU/L, TBII 7.9%, TSAb 122%, RAIU 7.2%) became euthyroid spontaneously. At the time of RAIU, the serum fT4 level was 1.2 ng/dL and TSH level was 0.22 mU/L, suggesting that the patient was at the recovery phase of PT. In another patient (TBII 0.1%, RAIU 10.4%), TBII remained negative, but the maintenance dose of MMI was continued for 8 years until remission. Third patient became euthyroid with a maintenance dose of MMI (10 mg/week), and TBII remained negative for 6 years, but relapse of GD was observed subsequently. The clinical course of these two patients suggested the possibility of mild form relapse of GD. In the other 3 TBII-positive patients, the TBII activity increased, requiring an increased MMI dosage, after the episode of thyrotoxicosis despite a marginally elevated RAIU (10%–15%), suggesting relapse of GD.
ATDs prescribed before the relapse of thyrotoxicosis (Table 2)| I) | Drug | Group A GD→PT |
Group B GD→Indefinite |
Group C GD→rGD |
| KI | 13 (37.1%) | 2 (33.3%) | 3 (5.1%) | |
| KI + MMI | 6 (17.1%) | 0 (0.0%) | 4 (6.8%) | |
| KI + PTU | 0 (0.0%) | 0 (0.0%) | 1 (1.7%) | |
| MMI | 10 (28.6%) | 4 (66.7%) | 39 (66.1%) | |
| PTU | 2 (5.7%) | 0 (0.0%) | 7 (11.9%) | |
| After tentative remission | 4 (11.4%) | 0 (0.0%) | 5 (8.5%) | |
| Total | 35 (100%) | 6 (100%) | 59 (100%) | |
| II) | Reduced dosage | 17 (48.6%) | 2 (33.3%) | 32 (54.2%) |
| Hypothyroid phase | 10 (28.6%) | 1 (16.7%) | 5 (8.5%) |
GD: Graves’ hyperthyroidism, PT: Painless thyroiditis, rGD: relapse of GD, KI: Potassium iodide, MMI: Methylmercaptoimidazole, PTU: Propylthiouracil, After tentative remission: Relapse of thyrotoxicosis diagnosed more than a year after the withdrawal of the antithyroid drug. Reduced dosage: Patients in whom the dosage of antithyroid drug was reduced before the relapse of thyrotoxicosis within one year. Hypothyroid phase: became hypothyroid after the relapse of thyrotoxicosis (GD or PT) with or without antithyroid drugs.
The ATD prescribed before the episode of relapse of thyrotoxicosis was mostly thionamide (78.0%) in Group C, but about 54.2% of the patients in Group A had been prescribed KI (Table 2-I). The incidence of PT among groups taking different ATD was 10 (18.9%) in 53 patients taking MMI alone, 2 (22.2%) in 9 patients taking PTU alone, 13 (72.2%) in 18 patients taking KI alone and 6 (54.5%) in 11 patients receiving combined therapy of KI and MMI or PTU. In about 11.4% of the patients in Group A and 8.5 % in Group C, relapse of thyrotoxicosis was observed more than 1 year after the withdrawal of the drug. In Group C, relapse of GD was observed when the ATD dosage was reduced in 54.2% of patients. Ironically, PT was also mostly observed when the ATD dosage was reduced in 48.6% of patients in Group A (Table 2-II). Episodes of hypothyroidism, suggested by the abnormally low serum fT4 level, were observed in 10 (28.6%) patients in Group A, mostly after increasing the dosage of ATD, which had been mistakenly administered without confirming RAIU (Figs. 1 and 2). In contrast, iatrogenic hypothyroidism during ATD treatment in relapsed GD was observed in only 5 (8.5%) patients in Group C (Table 2-II).
PT with positive TBII (PT on GD)There were 4 patients with PT (Group A) in whom TBII was positive when the RAIU was <5%/5 h (Fig. 4b). In a male patient, who had been treated with 100 mg KI and 5 mg MMI for 3 years, the serum fT4 and fT3 levels were markedly increased to 7.1 ng/dL and 10.6 pg/mL, respectively, with strongly positive TBII (61.7%), positive TSAb (228%) and a moderately enlarged thyroid gland (61 mL). He decided to be treated with radioactive iodine (131I) (RI). Both KI and MMI were withdrawn. However, the RAIU after strict iodine restriction for 7 days was 3.0 %/5 h. He was therefore diagnosed as PT on GD. RI therapy was cancelled, and he became spontaneously euthyroid after 1 month without ATD. He was well until he visited our hospital again 2.5 years later, when fT4 was 7.8 ng/dL, fT3 was 30 pg/mL, TBII was 99.0%, TSAb was 490%, estimated thyroid volume was 64 mL, and the RAIU was 68.6 %/5 h. He was diagnosed with relapse of GD this time and treated with 13.3 mCi 131I. He became hypothyroid after 2.5 years and was successfully treated with 75 μg L-T4, although both TBII and TSAb remained positive for 11 years and then became negative.
In the three other female patients with positive TBII in Group A, the serum TBII, TSAb and thyroid volume were 18.0%, 121% and 53 mL; 25.0%, 100% and 45 mL; and 28.2%, 138% and 24 mL. They became euthyroid within 2 months without ATD, suggesting PT, and TBII became negative after 0.5–2 years, resulting in remission of GD.
Relapse of GD with negative TBIIIn Group C (relapse of GD with a high RAIU), TBII was positive in only 66.1% of patients. TBII was negative in 20 (33.9%) patients (Fig. 4b) but became positive afterwards in 10 (50%) at 245 (111–623) days after the RAIU re-evaluation. These patients were suggested to be suffering from relapse of GD requiring ATD. A typical case is shown in Table 3. Both TBII and TSAb were negative at the time of relapse with a high RAIU (54.4%/5 h). TBII followed by a second-generation sensitive assay and TSAb became strongly positive four months later and became negative again after two to three years.
| Months after treatment | free T4 ng/dL |
free T3 pg/mL |
TBII IU/L |
TSAb % |
RAIU %/5 h |
Antithyroid Drug |
|---|---|---|---|---|---|---|
| –5 | 1.2 | 2.8 | 0.1 | 126 | none | |
| 0 (Relapse) | 4.7 | 18.5 | 0.1 | 133 | 54.4 | KI 100 mg |
| 0.5 | 2.2 | 9.0 | 1.8 | KI 100 mg | ||
| 1 | 2.3 | 10.4 | 233 | KI 100 mg | ||
| 3 | 1.5 | 8.8 | 35.6 | 955 | KI 100 mg | |
| 4 | 1.1 | 11.6 | 108.1 | 1,007 | KI 100 mg + MMI 5 mg | |
| 6 | 0.7 | 5.7 | 64.6 | 687 | KI 100 mg + MMI 5 mg | |
| 11 | 0.6 | 4.7 | 19.2 | 350 | KI 100 mg + MMI 5 mg | |
| 16 | 1.2 | 5.7 | 6.5 | 200 | KI 100 mg + MMI 5 mg | |
| 23 | 1.4 | 4.5 | 4.6 | 162 | KI 100 mg + MMI 7.5 mg | |
| 29 | 0.8 | 2.7 | 1.9 | 129 | KI 100 mg + MMI 5 mg | |
| 36 | 1.2 | 3.5 | 0.1 | KI 100 mg + MMI 2.5 mg | ||
| Reference value | 0.8–1.7 | 2.2–3.8 | <1 | <180 | 4–20 |
In this patient, the change in TBII was followed using a second-generation assay with the human recombinant TSH receptor (DYNOtest TRAb human kit, Yamasa Corporation, Chiba, Japan). TBII: TSH binding inhibitor immunoglobulin activity, TSAb: thyroid-stimulating antibody activity measuring c-AMP produced in FRTL-5 cells. RAIU: Thyroidal radioactivie iodine uptake. KI: Potassium iodide, MMI; Methylmercaptoimidazole, 0 (Relapse) : Values at the diagnosis of relapse of Graves’ hyperthyroidism.
Nine other patients were considered to have smoothly resolving hyperthyroidism, as they went into remission 802 (333–1,351) days after taking a small maintenance dose of ATD while TBII remained negative. The prognosis of one other patient was unclear.
The comparison of patients with negative TBII in Groups A and C (Table 4)| Number (%) | A) GD→PT 31 (60.8%) |
C) GD→rGD 20 (39.2%) |
p value | Odds ratio (95% CI) | |
|---|---|---|---|---|---|
| I) | Age (years) | 38.8 ± 17.4 | 42.3 ± 11.6 | 0.1273 | 0.9661 (0.9240–1.0102) |
| Sex (Male:Female) | 6:25 | 4:16 | 0.7021 | 0.7067 (0.1176–4.2478) | |
| TGHA (positive) | 26 (83.9%) | 7 (35.0%) | <0.0001 | 18.7860 (3.7391–96.4038) | |
| MCHA (positive) | 27 (87.1%) | 19 (95.0%) | 0.0783 | 0.1245 (0.0093–1.6679) | |
| II) | TBII (positive) | 0 (0.0%) | 0 (0.0%) | ||
| TSAb (positive) | 1 (3.2%) | 0 (0.0%) | |||
| RAIU (%/5 h) | 2.4 (0.9–3.1) | 27.6 (23.1–47.9) | |||
| III) | fT4 (ng/dL) | 3.2 (2.4–4.2) | 2.3 (1.9–3.7) | 0.4306 | 2.1488 (0.3398–13.5892) |
| fT3 (pg/mL) | 7.6 (5.6–11.6) | 5.6 (5.2–9.8) | 0.9017 | 0.9726 (0.6311–1.4990) | |
| fT3/fT4 (pg/mL/ng/dL) | 2.5 (2.0–2.9) | 2.9 (2.5–3.3) | 0.1719 | 0.2043 (0.0180–2.3131) | |
| Thyroid volume (mL) | 20 (10.6–25.4) | 18.4 (10–23.4) | 0.6533 | 0.9759 (0.8779–1.0848) | |
| Interval (days) | 1,876 (973–3,290) | 1,292 (1,084–2,098) | 0.3687 | 1.0003 (0.9996–1.0009) |
GD: Graves’ hyperthyroidism, PT: Painless thyroiditis, rGD: relapse of GD, TGHA, MCHA: anti-thyroglobulin and anti-thyroid microsomal antibody measured by hemagglutination, TBII: TSH binding inhibitor immunoglobulin, TSAb: thyroid-stimulating antibody, RAIU: thyroidal radioactive iodine uptake, Interval: between the date of RAIU before treatment of GD and the date of RAIU at the relapse of thyrotoxicosis. Odds ratio and 95% confidence interval (CI) are shown.
The clinical features of the 51 patients in Group A or C in whom TBII was negative at relapse were compared in Table 4. Thirty-one (60.8%) patients were in Group A (PT), and 20 (39.2%) patients were in Group C (relapse of GD), as shown by suppressed or high RAIU, respectively. A multivariate logistic regression analysis showed no significant difference in thyroid function test findings including the thyroid volume (Table 4-III), suggesting difficulty in differentiating relapsed GD and PT in TBII-negative patients with mild thyrotoxicosis without examining RAIU. The interval between the first episode of GD and the relapse of thyrotoxicosis seemed to be longer in Group A than in Group C. However, the difference was not significant. The only marked difference found was the higher prevalence of patients with positive TGHA in Group A than in Group C (Table 4-I), as was also seen in the analysis of all patients, including TBII-positive patients (Table 1-I).
The diagnostic accuracy of relapsed thyrotoxicosis without measuring RAIU (Table 5)| Clinical findings at relapse | Prevalence of painless thyroiditis | ||||||
|---|---|---|---|---|---|---|---|
| I) | All Patients at relapse | 35/100 | (35.0%) | p value | |||
| II) | A) | TSH Binding Inhibitor Immunoglobulin Activity (TBII) at relapse | |||||
| 1: TBII strongly positive | (≥30%) | 1a)/22 | (4.5%) | ||||
| 2: TBII slightly positive | (15%–30%) | 3/24 | (12.5%) | <0.0001 | |||
| 3: TBII Negative | (≤15%) | 31/54 | (57.4%) | ||||
| B) | Antithyroglobulin antibody (TGHA) | Positive | 30/60 | (50.0%) | 0.0006 | ||
| Negative | 5/40 | (12.5%) | |||||
| Antithyroid microsomal antibody (MCHA) | Positive | 31/88 | (35.2%) | 0.3650 | |||
| Negative | 4/12 | (33.9%) | |||||
| C) | 1: TBII(+) & TGHA (–) | 0/20 | (0.0%) | ||||
| 2: TBII(+) & TGHA (+) | 4/26 | (15.4%) | <0.0001 | ||||
| 3: TBII(–) & TGHA (–) | 5/20 | (25.0%) | |||||
| 4: TBII(–) & TGHA (+) | 26/34 | (76.5%) | |||||
| D) | Other factors suggesting intensity of Graves’ disease activity at relapse | ||||||
| 1: Large dosage antithyroid drugb) before relapse | 1/15 | (6.7%) | 0.0126 | ||||
| 2: Thyroid volume | Large | ≥60 mL | 2/10 | (20.0%) | |||
| Moderate | 30–60 mL | 14/30 | (46.7%) | 0.2597 | |||
| Small | ≤30 mL | 19/60 | (31.7%) | ||||
| III) | Early response during treatment | ||||||
| A) | Improvement of serum fT4 or TSH level without antithyroid drugc) | ||||||
| 1: Normalized or low serum fT4 level <2 months | 28/28 | (100%) | |||||
| 2: Normalized or elevated serum TSH level <4 months | 24/24 | (100%) | |||||
| B) | Hypothyroidism following thyrotoxicosis with or without antithyroid drugd) | ||||||
| All patients who became hypothyroid | 10/16 | (62.5%) | |||||
| 1: TBII Positive | 0/5 | (0%) | |||||
| 2: TBII Negative | 10/11 | (90.9%) | |||||
a) Painless thyroiditis on active Graves’ hyperthyroidism. b) Methylmercaptoimidazole ≥15 mg or propylthiouracil ≥150 mg with or without potassium iodide. c) In Group A (Painless thyroiditis), normalization of serum level was confirmed after 35 (26–56) days in fT4 (n = 34) and after 91 (37–126) days in TSH (n = 32). d) Hypothyroidism found within 3 months after the relapse of thyrotoxicosis during careful treatment. In Group A, hypothyroidism was diagnosed after 60 (55–71) days (n = 11) without drug.
The prevalence of PT as suggested by the intensity of Graves’ disease activity at the relapse of thyrotoxicosis during treatment of GD is shown in Table 5. The overall prevalence of PT was 35% (Table 5-I) and the prevalence was significantly higher (about 50%) when TBII was negative or TGHA was positive (Table 5-II). The prevalence of PT was 76.5% in the patients with negative TBII and positive TGHA. The prevalence of PT was only 4.5% if TBII was strongly positive (≥30%), and 6.7% if a large dosage of ATD had been administered before the relapse (≥15 mg MMI or ≥150 mg PTU). The difference in thyroid volume had only limited value for the differential diagnosis.
If the patients had been carefully followed without ATD after relapse, PT would have been strongly suggested if the serum fT4 and TSH levels normalized within 2 and 4 months, respectively, as was observed in 28 (80.0%) and 24 (68.6%) of the 35 patients in Group A, respectively (Table 5-III). If we had followed the patients for 3 months, with or without ATD, the possibility of PT would have been about 62.5% if they became hypothyroid, especially in TBII-negative cases (90.9%)
The diagnosis of PT was easily confirmed by the suppressed RAIU (<5%/5 h) at the relapse (Table 1, Figs. 3 and 4a).
The short- and long-term prognosis (Fig. 5, Table 6)
The non-remission curves determined using the Kaplan-Meier analysis after the relapse of thyrotoxicosis during treatments of Graves’ hyperthyroidism (GD). The non-remission rate during the five years after relapse is shown. Patients with painless thyroiditis (Group A) were followed without antithyroid drugs, except for 1 patient who was treated with 50 mg potassium iodide for 1 year. Remission was diagnosed when serum TSH and TSH receptor antibody levels normalized without recurrence of GD for one year.
| Group | Before treatment of GD | At relapse | Long-term prognosis of GD | Follow period | |||||
|---|---|---|---|---|---|---|---|---|---|
| TBII | n (%Total) |
TBII | n (%Total) |
Non-remission n %/n |
Remission n %/n |
Hypothyroid n %/n |
Total (years) |
Post-relapse (years) |
|
| A (GD→PT) |
(–) | 7 (20.0) | (–) | 31 (88.6) | 0 (0.0) | 26 (83.9) | 5 (16.1) | 10.6 (6.5–18.7) |
4.3 (1.8–9.8) |
| (+) | 28 (80.0) | (+) | 4 (11.4) | 1 (25.0) | 3 (75.0) | 0 (0.0) | |||
| Total | 35 (100) | Total | 35 (100) | 1 (2.9) | 29 (82.9) | 5 (14.3) | |||
| B (GD→Indefinite) |
(–) | 0 (0.0) | (–) | 3 (50.0) | 1 (33.3) | 2 (66.7) | 0 (0.0) | 17.2 (6.9–26.3) |
13.9 (4.0–21.9) |
| (+) | 6 (100.0) | (+) | 3 (50.0) | 2 (66.7) | 1 (33.3) | 0 (0.0) | |||
| Total | 6 (100) | Total | 6 (100) | 3 (50.0) | 3 (50.0) | 0 (0.0) | |||
| C (GD→rGD) |
(–) | 11 (18.6) | (–) | 20 (33.9) | 13 (65.0) | 6 (30.0) | 1 (5.0) | 19.5 (7.2–28.1) |
10.7 (3.1–21.1) |
| (+) | 48 (81.4) | (+) | 39 (66.1) | 27 (69.2) | 10 (25.6) | 2 (5.1) | |||
| Total | 59 (100) | Total | 59 (100) | 40 (67.8) | 16 (27.1) | 3 (5.1) | |||
| GD in general35) | (–) | 52 (9.5) | 19.5 (13.3–25.5) |
||||||
| (+) | 497 (90.5) | ||||||||
| Total | 549 (100) | 215 (39.2) | 301 (54.8) | 33 (6.0) | |||||
GD: Graves’ hyperthyroidism, PT: painless thyroiditis, rGD: Relapse of GD, TBII: TSH Binding Inhibitor Immunoglobulin activity, Non-remission: including ablated patients. Hypothyroid: Becoming spontaneously hypothyroid without ablation. The follow-up period was shorter in Group A than in other groups, although the difference was significant in the post-relapse perid (p = 0.0152) and not significant in the total follow-up period (p = 0.0778). Despite shorter follow-up period, long-term prognosis was significantly better in Group A than in Group C by the χ2 test (p < 0.0001). Regarding the differences between TBII (+) or TBII (–) patients, there was a siginificant difference in Group A (p < 0.0001), but no significant difference in Group C (p = 0.8062). Compared with the long-term prognosis of GD in general [35], the prognosis in Group A was far better, while that in Group C was markedly worse.
In Group A, ATD was discontinued after the diagnosis of PT was confirmed by the suppressed RAIU, except for in 1 patient in a very stressful condition who wanted to continue taking 50 mg KI for 1 year. The short-term prognosis at 5 years was far better in Group A than in Group C (Fig. 5). The time required for remission was also significantly shorter in Group A than in Group C according to the long-rank test (p < 0.0001).
Regarding the long-term prognosis, remission was observed in 82.9% while spontaneous hypothyroidism was observed in 14.3% in Group A (Table 6), rates that were much higher than those of 54.8% and 6.0% in GD in general, respectively [35]. A further analysis of the prognosis depending on the positivity of TBII suggested that the prognosis was especially good when TBII was negative in Group A. In contrast, in Group C, 67.8% of the patients were in non-remission despite the follow-up period being longer than in Group A, and 14 (23.7%) of the patients were treated with ablation 6.5 (4.6–13.6) years after the initiation of the treatment of GD. The non-remission rate in Group C (67.8%) was much higher than the rate of 39.2% in GD in general [35]. The presence of positive TBII at relapse had no significant influence on the long-term prognosis in Group C (Table 6). In Group B, remission was observed in 50.0% of patients, mostly in those in whom TBII was negative.
Clinical features of untreated PT or GD (Table 7)| Painless thyroiditis | Graves’ hyperthyroidism | p value | |
|---|---|---|---|
| Number | 105 | 1,124 | |
| Age (years) | 39.9 ± 14.8 | 36.9 ± 14.3 | 0.6366 |
| Sex (Male:Female) | 18:87 | 227:897 | 0.4301 |
| fT4 (ng/dL) | 2.5 (2.1–3.8) | 5.5 (3.9–8.5) | 0.0197 |
| fT3 (pg/mL) | 6.3 (5.0–8.2) | 14.5 (9.9–21.0) | <0.0001 |
| fT3/fT4 (pg/mL/ng/dL) | 2.4 (2.0–2.9) | 2.5 (2.0–3.1) | 0.3788 |
| TGHA (positive) | 25 (24.0%) | 392 (35.0%) | 0.5666 |
| MCHA (positive) | 52 (50.0%) | 854 (76.0%) | 0.0021 |
| Thyroid volune (mL) | 25 (15–29) | 29 (21–41) | 0.3978 |
| TBII (%) | 2.5 (0.1–5.5) | 40.0 (24.0–63.0) | |
| TBII positive, n (%) | 4/105 (3.8%) | 976/1,124 (86.8%) | |
| TSAb (%) | 113 (99–126) | 240 (142–478) | |
| TSAb positive, n (%) | 1/84 (1.2%) | 673/1,053 (63.9%) | |
| RAIU (%/5 h) | 2.4 (2.1–2.9) | 55.2 (36.9–70.5) |
A multivariate logistic analysis was performed excluding the factors that differed markedly from the diagnosis criteria, such as TBII, TSAb or RAIU. See legend for Table 1. In painless thyroiditis, TBII was marginally positive (16%–17%) in 4 patients and TSAb was marginally positive (190%) in 1 of 84 patients measured.
To evaluate the role of TGHA in PT, the clinical features of 105 patients with untreated low-uptake PT and 1,124 patients with high-uptake GD who visited our hospital during the same period were compared as shown in Table 7. Before a sensitive TBII or TSAb assay was available, the RAIU was examined in most patients with untreated thyrotoxicosis in our hospital, except for pregnant or lactating women. Among 1,124 patients with untreated GD (RAIU >15%/5 h), TBII was negative in 148 (13.2%), and both TBII and TSAb were negative in 87 (7.7%). Among the 105 patients with untreated PT (RAIU <5%/5 h), TBII was negative in most, being marginally positive (16%–17%) only in 4 patients. TSAb was positive in 673 (63.9%) of 1,053 patients in GD and marginally positive (190 %) only in 1 (1.2%) of 84 patients in PT. In this comparison, the prevalence of patients with positive TGHA or MCHA was lower in PT cases than in GD cases, although the difference was significant only for MCHA. Serum fT4 and fT3 levels were significantly higher in GD than in PT, but there was no significant difference in the fT3/fT4 ratio between the groups.
Both GD and Hashimoto thyroiditis are considered autoimmune thyroid diseases with antithyroid autoantibodies. The major difference is the presence of thyroid-stimulating TRAb in GD, resulting in hyperthyroidism. Although the mechanism underlying the development of PT is not clear, destructive changes in the thyroid gland with the follicular disruptions [10] are suggested to be responsible for the transient thyrotoxicosis without hyperthyroidism followed by hypothyroidism [2-10]. However, it must be pointed out that this so-called destructive thyrotoxicosis must be accompanied by an enhanced proteolytic process of Tg, as the destructive process of the thyroid alone, which is usually found in thyroid cancer, does not induce thyrotoxicosis, even when the serum Tg level is extremely high. Some inflammatory processes in the thyroid may also be present in PT, as found in subacute thyroiditis or subacute thyroiditis-like syndrome (STLS) seen in acute exacerbation of chronic thyroiditis [36] or amyloidosis [37]. Our study suggested that such PT-like reactions in the thyroid may occur frequently, even in the clinical course of GD.
If PT occurs during ATD treatment of GD, the diagnosis is difficult. Ironically, PT occurred frequently when the ATD dosage was reduced (Table 2-II). The patients then became hypothyroid after the ATD dosage was rapidly increased if the possibility of PT was overlooked (Figs. 1 and 2). Autoregulatory mechanisms in the thyroid gland are well known, and thyroid hormone synthesis was thought to be regulated by organified iodine compound X, probably an iodoaldehyde and/or iodolactone, which requires active thyroid peroxidase (TPO) to be synthesized, although the mechanisms underlying the effects of compound X remain elusive [38]. It is therefore plausible that a decreased ATD dosage might increase TPO activity and thereby increase the production of compound X, which suppresses the thyroid function, including that of sodium-iodide symporter (NIS), resulting in a decreased iodine uptake possibly accompanying Tg proteolysis and thyroid hormone release. From a therapeutic perspective, it is very important to keep in mind that PT can occur during ATD treatment of GD, especially when the dosage is reduced.
Although TRAb may be a useful marker for differentiating PT from GD, TBII was positive in 11.4% of PT patients and negative in 33.9% of relapsed patients with GD in this study (Fig. 4b). TBII measured by a first-generation assay and TSAb were reported to be positive in 15.1% and 18.9% of PT patients, respectively [39]. In our study, TBII and TSAb were positive in 3.8% and 1.2% of patients with untreated PT, respectively (Table 7), and in 11.4% and 5.7% of the patients with PT following GD, respectively (Table 1). Furthermore, in the relapsed patients with high-uptake GD, TBII and TSAb were negative in 33.9% and 64.4% of patients, respectively. This discrepancy might be due to issues with the sensitivity of the older assays used. However, the significantly higher prevalence of TBII-negative patients with relapsed GD than with untreated GD suggested that a repeated evaluation of the serum fT4 level during treatment might reveal early relapse of GD before the increase in TRAb activity in the serum, as shown in Table 3.
An interesting finding was the faster occurrence (Fig. 5) and the higher incidence of remission (Table 6) of GD in Group A than in Group C. Overt thyroid dysfunction, including PT, during immune checkpoint inhibitor treatment in cancer patients was associated with an improved progression-free survival and increased overall survival [26-28]. Our study suggested that the occurrence of PT during the treatment of GD might also be accompanied by immunological perturbation, suggesting a favorable prognosis resulting in remission. Destructive damage, including lymphocyte accumulation in the thyroid, may be associated with the mechanism of remission in GD [40]. The prevalence of spontaneous hypothyroidism, which was found in about 6% of GD patients in general after long-term follow-up [33, 35], was also as high as 16.1% among GD patients suffering from PT during the clinical course (Table 6). Except for patients with blocking-type TBII, spontaneous hypothyroidism is considered a clinical manifestation of end-stage chronic thyroiditis, suggesting the co-existence of GD and chronic thyroiditis over the long clinical course in the same patient [1, 19, 40]. The thyroid function of the patient may depend on the balance between the stimulating activity of TBII and preserved responsiveness of the thyroid gland itself. Therefore, it must be emphasized that if TBII is strongly positive even when low-uptake PT occurs, which may be called “PT on active GD”, relapse of GD can occur after several months when the resolution of PT occurs, as also reported by Momotani in postpartum patients with GD [16].
As shown in Table 2-I, PT was frequently observed during KI treatment. In Group A, 19 (54.3%) patients were treated by KI alone or KI and MMI before the episode of PT. Given the effect of excess iodide on the morphological changes in the thyroid [41, 42], KI treatment may precipitate the “iodide thyroiditis” reported by Edmunds in 1955 [43]. In the same year as Gluck reported convincing cases with PT [2], Savoie reported 10 cases of iodine-induced thyrotoxicosis in apparently normal thyroid glands, ranging from 1 to 40 months after exposure to excess iodine [44]. They all showed a typical clinical course of PT with a low RAIU followed by hypothyroidism.
Although Jod-Basedow or iodine-induced hyperthyroidism is well-known risk of administering iodine for endemic goiter [45], iodine-induced hyperthyroidism [46, 47], exacerbation of Hashimoto thyroiditis with reversible hypothyroidism [48] and amiodarone-associated thyroid dysfunction [49] have been reported in iodine sufficient areas. These functional abnormalities, in addition to pathological abnormalities [41, 42, 48], including a toxic effect observed on electron microscopy [50, 51], suggested the possibility of slow chemical ablation by excess iodide in susceptible patients. This effect may be one of the reasons for the increased incidence of remission or spontaneous hypothyroidism observed following an episode of PT (Table 6).
In general GD, a 54.8% remission rate was observed after long-term follow-up in our hospital [35]. Only a 27.1% remission rate seen in Group C (Table 6), suggested a roughly 50% lower chance of remission once relapse occurred during careful ATD treatment.
The high incidence of patients with positive TGHA in Group A was unexpected but interesting (Table 1-I, Table 5-II), as a significant association of TGHA at baseline with the development of thyroid dysfunction after nivolumab treatment was reported [25]. However, in the untreated patients with PT who visited our hospital during the same period (Table 7), the prevalence of the patients with positive TGHA was lower than that of those with untreated GD. The discrepant findings of a reduced prevalence of patients with positive TGHA among untreated PT patients compared with PT following GD suggested the possibility of non-autoimmune mechanisms for untreated so-called PT, as seen in painless subacute thyroiditis [52], STLS [37] or thyroiditis induced by excess iodide [44]. TGHA may be associated with the mechanisms involved in exacerbation of immune-mediated PT, as seen in PT following GD or nivolumab treatment. Downregulation by colloidal Tg on the TSH-stimulated process including the transcription of Tg, NIS or enzymes employed in Tg iodination has been suggested [53]. Antibody to Tg may affect this regulatory system in autoimmune thyroiditis. Recently, Kamijo reported 11 patients with PT following KI treatment for GD. Anti-Tg antibody was positive in 10 of the 11 patients [54]. The role of antithyroid antibody and iodine in the clinical course of GD or PT should be re-evaluated in the future.
In this study, repeated relapse of PT was suggested in 9 (25.7%) of 35 patients in Group A, including 4 patients who showed relapse more than 5 times. As the recurrence of post-partum PT following each delivery, the recurrence of non-postpartum PT has been reported relatively frequently [55-61], requiring RI therapy [57, 59] or thyroidectomy [58, 60] in some cases. Stressful events, as was suggested in Patient 1 in this study, or allergies to certain environmental factors [56] may be involved in the pathogenesis of non-postpartum PT.
The present study was limited by the use of a first-generation assay for TBII or TSAb in the main study. A second-generation assay was used to reconfirm the negative TBII value in the patient with high-uptake relapsed GD (Table 3) and to quantitatively follow the changes in TBII. The findings in that patient underscored the limited value of measuring TBII when the activity is low, even when using a quantitative sensitive assay, although the possibility of PT is less likely when TBII is strongly positive (Table 5-II). Regarding the concern about medical radiation exposure, radionuclide imaging and uptake studies used to play a significant role in the diagnosis of thyrotoxicosis. However, many previously appropriate indications may be no longer medically necessary after the introduction of ultrasonography and sensitive TRAb assay. They should be applied judiciously only in appropriate cases, such as imaging test for autonomous functioning thyroid nodule and uptake test before 131I therapy. Our study suggested that the repeated or unexpected ambiguous relapse of thyrotoxicosis during careful ATD treatment may be another possible indication of uptake test to evaluate the necessity to continue ATD treatment. The uptake test after appropriate iodine restriction is the golden standard to evaluate the responsiveness of the thyroid to TSH or TRAb [34]. Even if the result is indefinite, uptake value could show the real thyroid stimulating activity in TSH suppressed condition. Compared with 131I used in the past, 123I has far less radiation exposure risk without emitting β ray with short physical half-life of 13 hr. Absorbed dose after 3.7 MBq 123I administration was reported to be about 0.03 mGy in the general body including testis and ovary [62]. The cost is less expensive than the repeated TBII evaluations. It is important not to examine scintigraphy which increases the medical cost more than twofold or threefold. Since RAIU test of this study was performed about 20–30 years ago, when sensitive TBII assay was not popular, written consent was not obtained from the patients but all the patients understood the significance and accepted the RAIU examination.
It is important to recognize that PT or low-uptake thyrotoxicosis can often occur in the clinical course of GD and that the thyroid function reflects the balance between stimulating TRAb activity and the responsiveness of the thyroid tissue. Positive TBII does not rule out the possibility of PT (PT on GD) and negative TBII does not rule out the possibility of relapse of GD, suggesting that the repeated evaluation of the thyroid function aid in the early detection of hyperthyroidism due to a disrupted thyroid autoregulatory mechanism resulting in an enhanced RAIU or autostimulation before the detection of TBII in the serum (Table 3). When an unexpected increase in the serum fT4 level is observed during or after careful ATD treatment of GD, the possibility of PT is about 60% if TBII remains negative (Table 5-II). If the patients then become hypothyroid during treatment within 3 months, the possibility of PT increased to about 90% (Table 5-III). If patients lack serious signs and symptoms of thyrotoxicosis, including atrial fibrillation, it may be wise to carefully follow those without ATD. If the serum fT4 and TSH levels normalize within 2 and 4 months, respectively, the diagnosis of PT is extremely likely (Table 5-III). Otherwise, the readministration of ATD should be considered. The diagnosis can be confirmed by the suppressed RAIU (<5%/5 h) in the thyrotoxic state, which remains a valuable factor for differentiating PT from relapse of GD. Regarding the prognosis, the chance of remission of GD is high after an episode of PT unless TBII remains strongly positive, suggesting that PT may be one mechanism involved in remission, although frequent relapse of PT can occur.
The authors thank Dr. Brian Quinn for correcting the English. This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
None of the authors have any potential conflicts of interest associated with this research.
